‘A new star’ – but why just parenchyma for dendroclimatology?

Authors


(tel +49 (0)40 73962 452; email dieter.eckstein@uni-hamburg.de)

In this issue of New Phytologist, Olano et al. (pp. 486–495) present an amazing study on ray parenchyma of Spanish juniper (Juniperus thurifera) as a potential climate proxy. Here, this event is put into its developmental context.

‘At this point, Olano et al. come into play with their pioneering study on Spanish juniper, a long-lived evergreen tree species …’

Environmental information in trees is encoded in the year-to-year or even intra-annual variability of tree-ring width, earlywood and latewood width, wood density, cellular wood structure, and chemical composition of cell walls. By the mid-twentieth century, a great deal of such information had primarily been brought together on tree-ring width. Since then, a new quality of observations and interpretations of tree growth has gradually emerged through numerous technical progressions in microscopy, image analysis, topochemical and spectroscopical techniques, data acquisition and data processing. The driving force behind such development was the ambition of the scientific community to push the frontiers of tree-ring research forward by finding new variables in the annually formed wood and by realizing a time resolution of higher than 1 yr. As early as the 1920s, a higher time resolution had been achieved with the easy separation between earlywood and latewood width of conifers. In the 1960s, wood density profiles across the tree rings of conifers using X-ray densitometry became feasible, and around the same time, the focus started to be put on the cellular ring structure. In the meantime, these early approaches have been further improved and are increasingly applied in practice (e.g. Sass & Eckstein, 1995; Panyushkina et al., 2003; Rossi et al., 2003; Prislan et al., 2009; Seo et al., 2011).

In which direction could this process continue? The search for wood anatomical variables as environmental proxies, at its beginning largely unsystematic and erratic, can be traced back, at least, by half a century (e.g. Eckstein et al., 1976; Vaganov & Terskov, 1977). Soon after, Denne & Dodd (1981) had collected a vast quantity of literature concerned with environmental effects on cell dimensions, based on a variety of experimental results and on a diversity of interpretations. From the various cell types and functions in the xylem, the emphasis was, at first, on the water-conducting cells, tracheids and vessels, assuming that their dimensions were directly controlled by water supply. In fact, this was true in a few cases but it soon turned out that such a simplistic assumption could not be generalized: a number of more recent studies have confirmed rather the contrary, that is to say, the diameter of water-conducting cells may also be controlled by temperature, not only rainfall, and the key driver may come into action months before the tissue is formed, and not only during the few days, or at most, weeks when the tracheids or vessels, built by the cambium, differentiate and enlarge. The metabolic processes behind such delayed effects have not yet been fully explored experimentally. Meanwhile, a diversity of viewpoints concerned with the effects of changing growth conditions on the water-conducting cells (hydrosystem) have been published, so that Fonti et al. (2010) felt the necessity to distill some general emerging trends and perspectives for future research. One of such questions was made obvious by Schulte (2012) who considered not only the diameter but also the length of tracheids, as well as the number of pits per tracheid, and the diameter of the pit membranes along the ascending water column through the stem of a tree. In any case, concerted efforts involving not only wood anatomy and dendrochronology but also functional ecology and plant physiology are required.

Apart from the hydrosystem of trees, there are two further main tissues to be considered in view of their structural characteristics – the supporting and the storage tissue. All three tissue types are interconnected and extend throughout the entire body of a tree. The fact that the annually varying characteristics of the water-conducting cells are environmentally driven and thus reflect a balance between efficiency and safety of water transport for an optimal tree growth has been accepted as self-evident. For the supporting tissue, there is no such forceful justification for the annually changing dimensions of fibers/tracheids. The demand of a tree for mechanical support is not at first sight subject to annual or even intra-annual changes. Nevertheless, the annually varying amount of lignin (Gindl, 2001), controlling the compression strength, and the annually varying angle of cellulose microfibrils (Xu et al., 2012), controlling the tension strength, in both cases in the secondary walls of tracheids, were found to respond to climate. However, the underlying physiological mechanisms remain to be discovered by further in-depth ongoing experiments.

What about the storage tissue? For transporting, loading, storing and deloading of nonstructural carbohydrates in a tree, a system of parenchyma cells, both in the axial and radial direction, is involved. Although the quantity of material stored and transported may change over short time intervals, up to now it has been debatable whether the volume of the reservoirs and pathways changes accordingly. At this point, Olano et al. come into play with their pioneering study on Spanish juniper, a long-lived evergreen tree species, endemic to the western Mediterranean Basin. Xylogenesis starts in early May and ends in late October. The average annual rainfall in the area is 556 mm with an extended drought in July and August. Unlike tree-ring width, the annual amount of ray parenchyma, more precisely, the percentage of the cross-cut ring surface occupied by ray parenchyma, responded negatively to an above-average warm October in the year before growth, positively to an above-average wet May in the current year and negatively to an above-average warm August towards the end of the growing season. That is to say, after a cool end of the previous growing season, during a rainy period at the onset of earlywood formation, and during a cool time during latewood formation, the percentage of ray parenchyma increased, and vice versa. These results reflect statistically significant associations between varying amounts of ray parenchyma and climate variables. To understand the underlying physiological pathways, Olano et al. put forward a few assumptions for discussion.

However, before considering metabolic details, the variables measured should be scrutinized as to whether they are anatomically realistic and convincing. For this purpose, it is helpful to imagine or illustrate the three-dimensional network of the ray system throughout a tree ring and across a tree-ring border. It will, for example, show whether the onset of a ray (classified as NEWRAY by Olano et al.) is an offspring from a new ray initial or whether it is simply an already existing ray emerging from the subsurface of the cross-section to the surface. Such a three-dimensional illustration similarly could show why a ray disappears from a two-dimensional cross-section (classified as ENDRAY). According to Barghoorn (1940), uniseriate rays may decrease in height through the loss of one or more initial cells but this occurs very rarely; it is also possible that a fusiform initial elongates at both of its ends and thus squeezes between the ray initials and divides the ray spindle into two separate parts (Fig. 1).

Figure 1.

Radial section through the xylem, cambium and phloem of Taxus sp. (Left) Drop-out of one row of cells in a ray, both on the xylem side (A) and on the phloem side (A′) after the loss of one ray initial. (Right) Splitting of a ray by an apically growing fusiform initial (after Barghoorn, 1940).

Olano et al. did well to start with Spanish juniper, a conifer tree species without resin ducts and without ray tracheids, so that all rays are uniseriate and composed of parenchyma cells throughout – this is the most simple case within the vast structural and physiological diversity of the ray system of modern trees. The spirited approach of Olano et al. deserves acceptance and should be applied to other conifer species, at first, with a similarly simple ray structure for verification/falsification of the hitherto finding. Thereafter, further challenges await us; for example, the study of coniferous wood whose rays are composed of parenchyma cells in their central part for transport and storage of assimilates, and of one or more rows of radially oriented tracheids along the upper and lower spindle edge for radial water transport. The structural and physiological diversity of wood rays increases even further when broad-leaved tree species are taken into account; their rays generally are built of parenchyma cells throughout but their dimensions are widely varying (from a few cells to c. 0.5 mm in height and from uniseriate to c. 25 cells in width). In broad-leaved trees, the axial parenchyma, existing in much higher percentages than in conifers, also has to be taken into account as it is closely interlinked with the ray parenchyma, but at the same time with the hydrosystem. This is true uncharted territory in view of whole-tree functioning based on long-term genetic adaptation or on short-term functional acclimation (Fonti & Jansen, 2012).

Ancillary